5 - The Constitutionalization of Parties and Politics

Summary

Recent decades have seen a sharp rise in constitutional provisions regulating core aspects of democracy, including the rules about parties, voting, and elections. The trend is apparent in both democracies and nondemocracies, although democracies tend to constitutionalize slightly more matters. Constitutionalization can help democracy by tying the hands of politicians. Looking at cross-national data, we find that constitutionalizing democracy is correlated with higher levels of democracy. Constitutionalization of democracy carries advantages as well as risks. We illustrate the dynamics with short case studies of Kenya and Thailand.

5.1 Introduction

“War,” said Ross Perot, “has rules. Mud wrestling has rules. Politics has no rules.”¹ This view captures an increasingly common view of the political arena as one in which all bets are off, and any attempt to constrain the players is doomed to fail.

A moment’s thought, however, exposes how facile that view is, at least for democracies. The metaphor of politics as a contest that takes place in a public arena is an evocation of a repeated game, in which the rules are themselves constitutive of play. Democratic politics may be dirty, but it relies on a set of structures that provide for competition, and its continued maintenance depends on some consensus on these underlying rules.

While the idea that politics has rules is probably uncontroversial, few have noticed that, over the past decades, these rules are increasingly specified in the text of the constitution itself. This was not always the case. Constitutions used to be mostly silent on key issues like the regulation of political parties, voting rights, or the details of holding and administering elections. But over the past decades, many constitutions have come to regulate core aspects of the democratic process, including political parties, details of elections and the electoral system, and voting rights. To illustrate, according to our data, 83 percent of constitutions in force today regulate political parties, 71 percent give a court or electoral body the power to oversee elections, and 44 percent specify rules for electing the lower house. Indeed, we find a sharp rise in the constitutionalization of democracy in three related areas: (1) the regulation of parties, (2) rules relating to voting and direct democracy, and (3) rules relating to administering elections and their oversight.

This is a profound change, with potentially significant consequences. Constitutions are usually entrenched, meaning that they cannot be changed by ordinary democratic majorities, but instead require larger thresholds of legislative support for approval. Most constitutions also envision the practice of judicial review, meaning that courts can invalidate laws and regulations that contradict the constitution.² Constitutionalizing the rules relating to the democratic process, therefore, means that the rules of the democratic game will be harder to change and are subject to judicial oversight. This, in turn, implies a judicialization of democratic politics – core decisions about democracy are now made by constitutional and supreme courts.

Because of the nature of constitutions, the constitutionalization of democracy can aid democratic practices. After all, entrenching democratic rules and making them subject to judicial oversight means that ordinary majorities cannot tinker with them to create partisan advantage. Because of these qualities, constitutionalizing democracy can serve a hands-tying function; it allows constitution-makers to double down on their commitment to maintaining democracy.³ In addition, writing down specific rules can also provide clarity on the rules of the game.⁴ Such clarity is especially important when political conventions are weak, as is often the case during democratic transitions.

We find some evidence to support these ideas. We show how constitutional provisions protecting democracy served a clear hand-tying function in Kenya and that these same provisions appear to have helped prevent democratic backsliding. Looking at cross-national data, we find that constitutionalizing democracy is correlated with higher levels of democracy, which is consistent with the idea that constitutional rules on democracy can help protect democracy.

But we also add a note of caution. Constitutionalizing democracy gives an important role to high courts, in that they get to act as umpire over the rules of the democratic game. This might be a democracy-enhancing feature when courts are independent. But when courts are captured by the ruling coalition, they can interpret these same provisions in ways that are inconsistent with democracy.⁵ This risk is especially present for provisions that give substantial discretion to courts, such as bans on undemocratic parties. We illustrate this possibility with the case of Thailand, where the constitutional provisions banning certain kinds of parties has been deployed by the Constitutional Court to ban democratic parties. Overall, we conclude that, while promising, constitutionalizing the rules of democracy is not a panacea to prevent democratic erosion. Under some conditions, particularly when the military plays a strong role in politics, constitutional provisions may have the opposite effect from their stated purpose of protecting democracy.

5.2 The Constitutionalization of Democracy

One general trend in constitutional design over recent decades has been articulation: a growing propensity to regulate constitutionally with greater levels of specification. Constitutions written today tend to cover an ever-growing list of topics and deal with these topics in substantial detail. This trend has also been described as a move toward “constitutional codification,” or growing “constitutional specificity.”⁶ The trend toward constitutional articulation affects many different areas of constitutional law, including constitutional rights, judicial power, foreign policy, the separation of powers, and “fourth branch” institutions. Here, we explore the same with respect to some of the core ingredients of democracy: political parties, voting, and elections.

5.2.1 Data

To map and explore the constitutional articulation of core aspects of democracy, we draw on data from the Comparative Constitutions Project to select some two dozen variables relating to democracy. We only selected those that arguably reflect rules compatible with democracy. To illustrate, we do not include provisions that ban specific parties or declare a one-party state; even though these provisions surely deal with parties and elections. At the same time, we do include provisions relating to militant democracy that ban certain types of parties and empower certain institutions to ban them. While one might argue that such rules are undemocratic, their goal, at least in theory, is to protect democracy.⁷

A first set of variables relates to the role of political parties: whether (1) the constitution refers to political parties; (2) creates a right to form political parties; (3) bans certain types of political parties (but not specific parties); (4) whether the legislature is given the power to ban certain types of unconstitutional parties (but not specific parties); (5) whether the constitutional court, supreme court, or the judiciary is given the power to ban certain types of unconstitutional parties (but not specific parties); (6) whether an electoral court or electoral commission is given the power to ban certain types of unconstitutional parties (but not specific parties); (7) whether the constitution guarantees equality of political parties; and (8) whether the constitution specifies that political parties can initiate general legislation.

A second set of variables relate to voting and direct democracy: (9) whether the constitution makes claims of universal suffrage; (10) whether there are any restrictions placed on the right to vote; (11) whether the constitution makes voting mandatory; (12) whether the constitution prescribes electoral ballots ought to be secret; and (13) whether the constitution gives individuals the ability to propose legislative initiatives.

A third, related set of variables describes to the electoral systems and the mechanics of elections. They include: (14) whether the constitution establishes a voting threshold of a certain proportion of the votes for a party to be able to take a seat in parliament; (15) whether ordinary (constitutional or supreme) courts have the power to supervise election; (16) whether an electoral commission, electoral court, or both, exist to oversee the election; (17) whether the constitution makes arrangements for scheduling the elections; (18) whether a specialized body (and not the executive or legislature) establishes the shape and size of electoral districts; (19) whether the constitution has provisions on the public financing of campaigns; (20) whether the constitution has provisions setting limits on the money used for campaigns; (21) whether the constitution prescribes the election timing for executive and legislature (either same or different days); (22) whether the constitution specifies the electoral system for the lower house; and (23) whether the constitution specifies the electoral system for the upper house.

5.2.2 Trends

A cursory look at the data reveals a growing constitutionalization of core features of democracy. Figure 5.1 shows the average number of all twenty-three democracy provisions over time. It shows a gradual increase in the number of electoral provisions in national constitutions. In 1810, the average number of provisions was 0.3; today, it is 6.9. But of course, there is substantial variation across countries. Even today, some constitutions do not enumerate any provisions relating to democracy, such as the absolute monarchies of Saudi Arabia, Brunei, and the United Arab Emirates.

Figure 5.1 Number of democracy provisions on world map.

The US Constitution today includes only two democracy provisions – restrictions on abridging the right to vote under the Fifteenth, Nineteenth, and Twenty-Sixth amendments, which we code as being a claim to universal adult suffrage. But the US Constitution is notoriously silent on parties. In fact, the founders of the American republic sought to create a system of government that would protect liberty by retarding the formation of parties. As James Madison put it in Federalist 10, factions were a danger to popular government. These entities, composed of “some common impulse or passion, or of interest, adverse to the rights of other citizens,” were an evil to be avoided through careful institutional design.⁸ Despite Madison’s best intentions, however, parties emerged early in the Republic as useful mechanisms to coordinate behavior in the legislature. As American democracy evolved and expanded in the nineteenth century, the party system changed as well, but constitutional language did not keep up. The Constitution was amended to provide for nondiscrimination in the provision of rights to vote for racial minorities, and later women and youth, but parties remain absent from the text. The time, place, and manner of congressional elections remain in the hands of state legislatures, which, contrary to Madison’s expectations, have become hotbeds of partisan self-dealing. In other settings, such questions are increasingly taken out of the realm of partisan politics and placed into the constitution.

On the other end of the spectrum, Kenya, starting in 2010, is the country with the largest number of constitutional democracy provisions. Its constitution contains no fewer than fifteen such provisions. It is followed by Liberia and Thailand (from 2007–2012), which both have fourteen democracy provisions.

When depicting the same data on a world map, we can see some regional trends, with a notable concentration of a high number of such provisions in both Latin America and Africa, which are areas with a good deal of constitutional turnover.

We see the same trends if we look at each of these categories separately. Consider, first, political parties. At the turn of the twenthieth century, exactly two constitutions referred to political parties: those of Colombia and Greece.⁹ Today, 83 percent of the 193 national constitutions in force contain such a reference. As of 2020, 29 percent of constitutions prohibit certain parties or types of party programs, and some 13 percent provide that a court determine whether a political party is unconstitutional. These phenomena were unknown in 1900. Figure 5.2, Panel A shows the increase of provisions on political parties over time, a trend that accelerated after World War II.

Figure 5.2 Average number of provisions relating to parties, voting, and elections.

We observe the same basic trends for voting and direct democracy. Figure 5.2, Panel B depicts the average number of provisions relating to voting. It is worth noting that these provisions are older than the political party provisions. We see them start appearing in the 1800s and steadily increase in number over time.

Finally, we also see the same trend for features of the electoral systems. Figure 5.2, Panel C depicts the average number of these provisions and reveals how they have increased over time. One increasingly popular constitutional design choice is to establish some form of electoral commission. This is a distinct body (or several) established to manage tasks like boundary delimitation and election management. In 1900, only Colombia and Liechtenstein provided for an electoral commission or/and court to oversee elections, but today 71 percent of constitutions constitutionally establish such a commission or give courts the power to oversee elections. Another noteworthy trend is to clarify the electoral system for parliamentary elections. In 1900, the four constitutions that clearly specified a voting rule for the lower house had a simple version of plurality or majority rule. Because of significant innovation in the understanding and design of voting systems during the twentieth century, contemporary constitutions not only have to choose between majoritarian and proportional representation but will sometimes specify a remainder formula, identify special constituencies for particular groups, or create a complex mixed system involving both proportional representation and districts. The five constitutional articles describing Sri Lanka’s mixed method electoral system for the lower house are nearly as long as some entire national constitutions.¹⁰ They illustrate the tendency to constitutionalize the specifics of parliamentary elections.

5.2.3 Relationship with Democracy

Are these new constitutional features associated with regime type? The data reveals that these features can be found in democracies and autocracies alike, although they are somewhat more common in democracies.

We can observe this by analyzing the trends for democracies and autocracies separately. Figure 5.3 depicts the growing constitutionalization of democracy for autocratic and democratic regimes (we consider a country to be democratic if its polity2 democracy score is over 4 [on a scale from -10 to 10]). The dashed line denotes democracies; while the dotted denotes autocracies and solid captures the full sample. The first panel shows the trends for all democracy provisions, the second panel for parties, the third for voting, and the fourth for the electoral process. Overall, the graphs reveal that the trend toward constitutionalization is slightly more pronounced in democratic countries. By 2016, the average democracy constitutionally enumerated 7.7 of the variables we collected, while the average autocracy enumerated 6.5. But Figure 5.3 also reveals that this difference is mostly driven by constitutional provisions relating to the electoral process. By contrast, the second panel reveals that the constitutionalization of parties is just as common in democracies as autocracies.

Figure 5.3 Average number of provisions relating to parties, voting, and elections by regime type.

Is the constitutionalization of democracy associated with higher levels of democracy de facto? It is notoriously difficult to sort out causation with cross-national data, and what follows is merely an initial exploration of whether constitutional provisions on democracy correlate with democracy de facto.

To explore the relationship between de facto democracy and constitutional rules relating to democracy, we estimate a simple OLS panel regression with the well-known polity2 democracy scale as the dependent variable. The model includes country-fixed effects, year-fixed effects, and a linear time trend. With these, we control for non-time varying country characteristics, common trends, and global shocks. We also experiment with adding a standard set of control variables: GDP per capita, population size, and civil war. Robust standard errors clustered are clustered by country to account for serial correlation.

Table 5.1 reports results from this exercise. It reveals that correlation between constitutionally entrenched rules relating democracy and de facto democracy is positive and statistically significant. For all the models capturing all provisions (columns 1 and 2), adding one additional constitutional feature (out of twenty-three) on democracy is associated with an increase of 0.455 on the 21-point democracy scale. While we should not interpret these as causal relationships, they are suggestive that these constitutional features are associated with somewhat higher levels of democracy.

Table 5.1 Relationship with democracy

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
all provisions	0.455***	0.426***
	(0.100)	(0.122)
parties			0.585**	0.515**
			(0.252)	(0.259)
voting					0.925***	0.984**
					(0.333)	(0.439)
elections							0.707***	0.706***
							(0.180)	(0.225)
controls	no	yes	no	yes	no	yes	no	yes
country FE	yes	yes	yes	yes	yes	yes	yes	yes
year FE	yes	yes	yes	yes	yes	yes	yes	yes
time trend	yes	yes	yes	yes	yes	yes	yes	yes
Observations	12,849	8,712	12,858	8,721	12,893	8,756	12,889	8,752
R-squared	0.707	0.731	0.698	0.723	0.701	0.728	0.705	0.731

Note: the dependent variable is democracy. ***p < 0.01, **p < 0.05, *p < 0.1.

5.3 Why Constitutionalize Democracy? An Exploration

Hands-tying. What motivates drafters to regulate core aspects of democracy in the constitution? A likely reason is that constitutions are thought to serve as “pre-commitment devices.”¹¹ The key idea here is that, when certain rules or values are constitutionalized, they are placed outside of the reach of ordinary politics. Because constitutions are harder to change than ordinary laws, ordinary democratic majorities cannot change constitutional rules. Additionally, courts are usually empowered to enforce these rules and to invalidate laws that contradict them. This also means that when disputes over interpretation or application arise, courts can further clarify and enforce these rules. The combination of entrenched and justiciable rules makes these rules harder to undermine and allows drafters to tie the hands of future democratic majorities.

This hands-tying logic reflects a certain amount of distrust of those who will govern under the constitution. Such distrust is often present during democratic transitions, when democracy is not taken for granted and constitution-makers use the constitution to attempt to lock in their commitments to a liberal democratic order.¹²

Under such conditions, creating a new machinery to run elections and to oversee democratic competition will make good deal of sense. The constitutionalization of parties likewise can serve a hand-tying function. The idea that constitutions should regulate and protect political parties can be traced back at least to Hans Kelsen, who noted that it is crucial “to anchor political parties in the constitution and give legal form de facto to what they have long since become: organs forming the will of the state.”¹³ On the one hand, modern constitutions protect parties. They do so through a right to form political parties, which ensures that constitutional mechanisms are available to prevent existing players from monopolizing the machinery of politics. Party rights usually also offer protections for smaller parties, preventing the state from intervening in a party’s internal affairs or making party registration requirements too arduous.¹⁴ In some cases, it might also entail certain guarantees for opposition parties to be represented in the democratic process.¹⁵ Empirical evidence suggests that rights to form parties are among those rights that are most likely to be effectively enforced, in part because parties themselves use these rights to protect their interests.¹⁶

But constitutions do not only protect parties; they also regulate them. The most common approach is to ban parties that propagate certain viewpoints. The practice of requiring parties to be democratic became relatively common after World War II. The rise of Nazism through mechanisms of parliamentary democracy led to great angst, and in 1937, the German political scientist Karl Loewenstein coined the term “militant democracy.” Concerned with the inadequate democratic response to the rising threat of fascism, he called for a set of legislative and legal techniques that would allow democracy to defend itself against threats that emerge from within. “Constitutional scruples” he noted, “can no longer restrain from restrictions on democratic fundamentals, for the sake of ultimately preserving these very fundamentals.”¹⁷ Loewenstein went on to catalog techniques used by inter-war drafters that could constrain autocratic elements within society. But the prototype instrument of a militant democracy limits parties to those that are democratic. The design strategy of combining party rights with party bans, then, seems to reflect a simultaneous desire to recognize parties as important institutions of governance but also to set some boundaries on the ideological playing field.

Coordination. Another reason why constitutionalizing the rules of democracy can aid democratic practices is that constitutions can aid coordination around the rules of the game. If everyone has common understandings of the rules of politics, there is little additional advantage from writing them down. But there may be times when existing understandings break down, either because of exogenous change or because of the escalation of partisan competition. This can create incentives to clarify the rules of the game, so that parties can coordinate better in the future.

Coordination is especially important in the context of democratization. In such a context, new systems of elections and voting are often being set up from scratch. When starting from scratch, clarifying the basic rules of the game is particularly important as there are no preexisting political conventions to rely upon.

In a similar vein, coordination also becomes important when democratic political process become more open in character. The nineteenth century saw the rise of political parties around the world, be they the Colorados and Blancos in Uruguay, the Liberals and Farmers parties in Sweden, or the Liberals and Conservatives in Canada. What these party systems had in common at this time is that they are largely elite affairs, in which parties regulated their interactions through reciprocity. But industrialization brought demands for an expansion of the franchise. As mass-based parties emerged out of labor movements in the early twentieth century, there was a sharp rise in polarization, putting pressure on norms of reciprocity among parties. In such a context, it made sense to start clarifying the rules of the political game and the role of parties within it. As democratic politics became more open and polarized, having clear rules helped parties to coordinate and channel their competition.

The explanations of hands-tying and coordination are not mutually exclusive, and likely work together in specific cases. For example, in the context of democratization, constitutionally specifying rules of the democratic process can be motivated by a joint desire to pre-commit to multiparty democracy as well as to clarify basic rules. Indeed, it is likely no coincidence that we see the constitutionalization of democracy take off during the wave of constitution-making in the 1990s.

5.4 The Case of Kenya

We can see the importance of distrust of political elites in the constitution-making process undertaken in Kenya in the early 2000s, culminating in the Constitution of 2010. As noted above, this constitution is noteworthy because it has the largest number of provisions dealing with democracy of any constitution in force today. It is therefore worth probing the motivations behind this document and to explore how these provisions have been used in practice.

The Kenyan constitution-making process had followed an uneasy decade of democratization, during which longtime strongman Daniel Arap Moi retired in 2002. The prior text, adopted at independence in 1963, included an Electoral Commission, but had no mention of parties, and left the rules about voting to ordinary legislation. This document governed a rigged political process, and the Electoral Commission had no financial autonomy. Mass action in 1997 by civil society led to a constitutional amendment increasing the number of electoral commissioners and giving political parties more say in their selection. In addition, a Constitution of Kenya Review Act sought to “facilitate the comprehensive review of the Constitution by the people of Kenya.”¹⁸ This required the creation of the Constitution Review Commission (CKRC) to consult broadly and produce a draft. The draft would then be debated at a National Constitution Conference (NCC), made up of politicians and civil society, before going to parliament and the public for final approval.

Headed by famous civil society activist and scholar Yash Pal Ghai, the CKRC produced a draft after wide public participation. A summary of the public views presented to the CKRC included various criticisms of political parties: there were too many of them, they were too autocratic, and without internal democratic procedures; they were vehicles for self-interest; they engaged in violence and hooliganism; and they did not seek to advance the national interest.¹⁹ The solution, according to the report of the CKRC, was that “Political parties, as institutions of democratic and republican governance … should be regarded as constitutional organs that should be provided for in and regulated by the Constitution.”²⁰ Accordingly, the draft finally approved by the NCC in 2004 (known as the Bomas draft after the location of the NCC meetings) contained a major section on political parties, with rules, regulations, and provision for a code of conduct, as well as a fund for campaign expenditures, which included limits on party activities. Ghai and his colleagues sought to limit parties’ ability to engage in corruption and to steer their activities toward democratic representation. In addition, the draft also fixed a time for parliamentary elections and introduced a clear right to vote.

However, as the process had been designed, the parliament had a chance to modify the draft before it went to public referendum for approval. Attorney General Amos Wako, a consummate political insider, was able to take control and watered-down key provisions in a process negotiated by the governing political parties without the opposition. The political-parties section shrank from ten articles to just two. The executive branch was made stronger. Civil society was outraged at this perceived capture of the draft by politicians, and the Wako version was rejected by public referendum in 2005 by a resounding vote of 58 to 42 percent. Subsequent elections were marred by significant violence, reflecting the largely tribal organization of Kenya’s existing political parties. The country stood on the brink of unraveling, and the international community came in to broker a reset.

A second constitution-making process followed, in which initial drafting was done by a nonpartisan Committee of Experts, which included three foreigners. This body was charged with drawing the best of the earlier proposals into a new harmonized draft, which would presumably reflect a consensus. Its members were well aware of the need to reflect both public demand and elite consensus in this process.²¹ The dynamics were similar to the prior episode, in which the Committee proposed a draft with an extensive section on political parties, but this was then submitted to a Parliamentary Select Committee, which reduced the number of articles on parties from nine to two. The draft then went back to the Committee of Experts for presentation to the public. The draft was ultimately approved by a referendum in August 2010, with more than 2/3 of voters supporting adoption, in a context of great distrust of politicians.

The resulting document included several global innovations, including a whole chapter on leadership and integrity, and established good governance, integrity, transparency, and accountability as national values. More specifically, the document creates a body to manage elections and oversee politics – an Independent Electoral and Boundaries Commission – and requires public participation in the process of delimitation of districts.

Even though the number of provisions regulating parties were reduced by politicians, the final document regulates parties substantially. Notably, the legislature is instructed to enact legislation regulating political parties, and Article 91 contains myriad constitutional rules about parties. These include requirements that parties have a national character and that they not be founded on ethnic, linguistic, regional, or other lines; that they have internal democracy in terms of their governing bodies and leadership; that they abide by principles of good governance; respect human rights, including those of minorities and marginalized groups to participate in the process; observe a code of conduct for parties; and several others.²²

Overall, these new provisions can be best understood as efforts to tie the hands of politicians on behalf of the public, to constrain and channel political activity in beneficial ways. For parties, the provisions allowed for better coordination among themselves on the rules of the game: knowing that the rules were present, parties had some incentive to “play fair” under the new framework, rather than escalating competition in ways that violated the terms of the constitutional text.

Did the 2010 Constitution’s strategy of “spelling things out” strategy work? Although Kenya’s political class is among the world’s most venal – the first act of parliament after approving the draft of new Constitution was to vote a pay raise for members – the text seems to have had an impact on the political system. A recent attempt by then-President Uhuru Kenyatta and his former rival Raila Odinga to cartelize the political system by unilaterally creating seventy new parliamentary constituencies and a new prime minister position was challenged by voters and minor parties. Among other provisions, the proposal – known as the Building Bridges Initiative – restricted judicial review of the boundary delimitation process and utilized a provision on popular initiative to propose the reforms. The courts rejected the attempt as procedurally flawed and incompatible with the Constitution, thereby foiling a partisan lock-up.²³ The High Court judgment in particular emphasized the role of the scheme in undermining the political process and the independence of the Commission, thus infringing on the rights of smaller parties and undermining the basic scheme of the Constitution. In August 2022, Odinga lost the presidential election to William Ruto, and outgoing incumbent Kenyatta peacefully turned over power to the latter – the first orderly alternation under the new charter.

To conclude, the Kenya story illustrates the democratic politics of constitutionalization: pressures for greater democratization led to the introduction of new ideas for regulation and strengthening constitutional protections for regulators. Both dimensions of parties and electoral mechanics were constitutionalized. The actual process took multiple iterations and involved civil society, technocrats, and sitting politicians, all under the supervision of courts at the phase of implementation. A major attempt to capture the process by two large parties, begun in 2020, was decisively rejected by the courts, using the constitutional text. In short, Kenya’s democracy is far from perfect but no doubt in much better shape than when Moi stepped down two decades ago, and the constitutionalization of democracy helped to secure this.

5.5 A Note of Caution

One important implication of constitutionalizing core features of democracy is that it empowers courts to police the boundaries of democratic politics. This is by design: when democracy is constitutionalized, constitution-makers envision that a neutral umpire can enforce constitutional boundaries in the face of partisan conflicts. Thus, high courts can now call out violations and clarify the rules of the democratic game when there is ambiguity on what these rules require.

When courts are independent, constitutionalizing the rules of democracy might make a whole lot of sense to protect democracy. But the central role for the court also comes with a vulnerability, which is that these same constitutional provisions can be used to accomplish partisan goals when courts are not independent. When the rules of democratic politics are constitutional in nature, then controlling a court becomes a powerful tool that allows would-be autocrats to stack the deck in their favor. Not only can courts reinterpret constitutional rules on democracy, their rulings enjoy priority over ordinary laws, which means that ordinary democratic majorities cannot simply alter them.

The pay-off of controlling the court might be highest for the provisions that represent broad and flexible standards. It is more difficult (though not impossible) to reinterpret clear rules, such as the day on which elections are to be held.²⁴ But other provisions give more discretion to courts. The ban on undemocratic parties is a prime example: it leaves to the courts to judge which policies and programs are undemocratic. When courts are captured, constitutional bans on undemocratic parties can become a tool to lock in current configurations. For example, in 2017, the authoritarian government of Cambodia had the Supreme Court disband the country’s main opposition group, the National Rescue Party, leaving strongman Hun Sen in control for an upcoming election. Individual members of the party were subjected to a five-year ban from running for office, as Hun Sen is reputed to be grooming his son to succeed him.

Thus, when courts become central to the functioning of democratic politics, it raises the stakes of controlling them. In the recent wave of democratic erosion, one of the first moves of would-be authoritarians is to control courts. When core features of democracy are constitutionalized, this only adds to the incentives to control courts.

Another factor that may lead to abuse of these provisions is a powerful role for the military in politics. Militaries often see themselves as the guardians of a set of fundamental values in the constitutional order, be they secularism (as in Turkey), religion and property (as in Chile under Pinochet), or the territorial integrity of the country (as in Pakistan or Nigeria). This self-conception can lead them to closely monitor civilian politics, even in a democratic era. They often align with judiciaries, with whom they share a hierarchical organization and technocratic orientation.²⁵

There is a long history of military-aligned governments seeking to use the machinery of party bans to maintain power and shape politics. Perhaps the paradigm case is Turkey, which banned Islamist parties several times before the eventual success of Recep Tayip Erdogan. The case of Thailand illustrates how provisions constitutionalizing democracy can be a double-edged sword. As in Kenya, the Thai Constitutions of 1997 and 2007 contain a large number of provisions regulating democracy. But in the Thai case, the tools of constitutional regulation have been used to constrain democracy itself on behalf of military-bureaucratic institutions that are at the core of the political system.

Thailand is a country with a long history of alternating between corrupt civilian governments and military regimes, and in 1997, a new constitution was adopted to try to end this cycle. Known as the “People’s Constitution” because it came out of a popularly elected constitutional assembly, the 1997 Constitution was the most democratic in Thai history. It contained numerous rights, political reforms to facilitate electoral majorities, and a network of “post-political” institutions to constrain and clean up democratic politics. Central here was a new Constitutional Court. Under Section 63 of the Constitution, the Court was given the power to disband any party found to be seeking to gain power by illegal means or to overthrow the “democratic regime of government with the King as Head of State.” This is an old formula in Thailand, which seeks to balance the formal requirements of democracy with a semi-divine monarchy.²⁶ We do not know who pushed for the inclusion of these provisions, but it likely reflected a combination of those with a genuine motive to protect democracy along with elements of the conservative bureaucratic-military elite.

Populist billionaire Thaksin Shinawatra dominated subsequent elections and held the premiership from 2001 to 2006, but conservative forces resisted him. Polarization increased sharply, and by the mid-2000s competing political factions were engaged in open and violent struggle in the streets of Bangkok. The Constitutional Court had dismissed corruption charges against Thaksin to allow him to take power in 2001, and sought to maintain a neutral role, but the rising political turmoil led the country’s revered monarch to demand that courts address the problems. In 2006, a military coup sent Thaksin into exile. The Constitutional Court subsequently took up a case against Thaksin’s Thai Rak Thai party, eventually deciding that it was banned under Section 63 of the Constitution. Polarization had drawn the courts into one side of a partisan divide.

The 2007 Constitution, produced by technocrats working with the military government, drew on the basic structures of 1997, with multiple regulatory institutions to govern elections, and contained an elaborately specified electoral system. It retained much of the language of the 1997 constitution with regard to parties, including the power of the Constitutional Court to ban parties under Section 68.²⁷ Section 65 of the 2007 document added new language, worth quoting in full:

A person shall enjoy the liberty to unite and form a political party for the purpose of making political will of the people and carrying out political activities in fulfillment of such will through the democratic regime of government with the King as Head of the State as provided in this Constitution.

The internal organization, management and regulations of a political party shall be consistent with fundamental principles of the democratic regime of government with the King as Head of the State.

Members of the House of Representatives who are members of a political party, members of the Executive Committee of a political party, or members of a political party, as to the number prescribed by the organic law on political parties shall, if of the opinion that their political party’s resolution or regulation on any matter is contrary to the status and performance of duties of a member of the House of Representatives under this Constitution, or contrary to or inconsistent with fundamental principles of the democratic regime of government with the King as Head of the State, have the rights to refer it to the Constitutional Court for decision thereon.

The motive for this section seems to reflect a deep distrust of political parties on the part of military technocrats, turning party members into monitors. When a member finds that the party leadership is acting outside the bounds of Thailand’s democratic system, they can refer the case to the Constitutional Court. Section 68 goes on to provide the Constitutional Court with the power to disband unconstitutional parties for actions incompatible with the democratic regime with the King as Head of State. More broadly, the courts generally serve as guardians to limit the domain of politics, with power to strike resolutions of parties and channel their activities. One might describe this section as abusive from the outset – an attempt on the part of military leaders to control democratic politics.

The 2007 Constitution quickly became an instrument to limit democratic politics. Despite having left the country after his prior party was banned, Thaksin Shinawatra did not disappear as a political force. His allies reorganized Thai Rak Thai into a new People’s Power Party, but this too was banned by the Constitutional Court in December 2008. Another successor party, Pheu Thai, won subsequent elections, and Thaksin’s sister Yingluck took over as prime minister. Conflict continued, occasionally erupting in violence. While the courts had in the early years of Thaksin’s ascendance sought to play a neutral role, they gradually took on the task of enforcing the demands of Thailand’s traditional conservative elite against populist challengers. By 2014, a new coup sought to restrict the domain of elections and parties even further. Thaksin remained out of the country and seemed to have been defeated through the efforts of multiple party bans along with two military coups.

This meant that there was something of an opportunity for a fresh start in Thai politics. When the new Constitution of 2017 was adopted, several new parties formed to compete. A leading challenger was the Future Forward Party, which represented urban and young voters, and was led by a scion of a billionaire family, Thanathorn Juangroongruangkit. When this party proved to be fairly popular, however, it came to be seen as a threat to the party led by coup leader Prayuth Chan-Ocha. On relatively thin evidence, the Thai Constitutional Court found that Thanathorn had violated campaign laws and removed him from his seat in parliament in 2019. The next year, the Future Forward Party was disbanded by the Court. Its successor, the Move Forward Party that won an electoral victory in 2023, is currently facing a number of complaints asking for its dissolution.

The Thai story illustrates that constitutionalization of parties can be a two-edged sword and puts a great deal of faith in those who will make determinations of compliance with core norms. Even if adopted to protect democracy initially, constitutional provisions can be reinterpreted and repurposed. Party bans raise the stakes of controlling a constitutional adjudicator and increase the temptation to abuse the power. As always, when the institutions charged with protecting democracy become aligned with one faction in local politics, there is great potential for abuse.

5.6 Conclusion

Whether constitutionalizing democracy will serve a genuine hands-tying function or instead be subject to abuse will depend on context, and this can change over time within a single constitutional order.

In the case of Kenya, the ability of the courts to rely on constitutional principles and procedures was central to rebuffing an attempt by the two largest parties to cartelize the political system and to facilitating a transfer of presidential power in 2022. The constitutionalization of elections and parties has led to a vigorous jurisprudence, which simply would not have been possible absent the rules laid out in the 2010 Constitution . The broader positive correlation between constitutional regulation and measures of de facto democracy gives us reason to believe that this story is not confined to Kenya but that we might see similar dynamics at work in other cases such as Liberia and Nepal, which both have high levels of regulation.

However, constitutionalization is not a panacea. As our Thailand case study illustrates, the presence of provisions on elections, voting, and parties in a constitution can in practice be abused. In the wrong hands, constitutionalization might make things substantially worse for democracy. Hard limits on the political process are themselves a double-edged sword and may be subject to abuse by military-aligned governments in particular. This implies that, when considering whether or not to include constitutional provisions on the political process, democrats have to focus on who will be interpreting them down the road, and to choose the provisions least likely to be abused